Presently, every company has access to huge sets of information and high-capacity processing power. Simultaneously, tools are widely spread and capable of answer multiple purposes in complex environments. Specifically for forecasting, there are multiple software prepared with several algorithms, pre-processing methods and advanced UI – nowadays, the challenge of having a forecasting module rely more on reliance, adaptability and deployment.

As a result, the difficulty falls on setup and adaptation to business requirements. More important than connecting the plug is knowing how to exploit and adapt all the modules provided. Faced with skeptical teams and poor results, successfully deploying a well-establish forecasting tool has been tough challenge.


To improve the results and foster team’s confidence, the approach followed 3 key structures:

System interpretability – Explore and translate each module and key forecasting algorithm to an application-based framework focused on actionable use-cases that helps the user to understand and review each outcome.

Accuracy improvement – Leverage and adjust all features to be tailored by business needs (disable, adapt and develop new modules). Subsequently, adequate the parameters for each set of products.

Simulation & Monitorization – Use advance simulation to compare business KPIs between the old and the new process and carefully supervise forecasting accuracy over time.


To foster confidence and expertise within teams, using advanced simulation modules and explaining the concepts of each forecasting module were critical lever to success. Consequently, an ambitious rollout plan was efficiently conducted with full support and comfort each operational team.

By bridging the gap between technical and business expertise was possible to adapt the new forecasting tool. This allows the system to be more responsive to business specificities and adjusted to the features of each product category.

Beside gains on processes and interpretability, the results showed more than 10p.p. improvement on accuracy and a more stable forecast compared with the initial design.

Related Case Studies



Business analytics involves several methods and tools that can be organized into three dimensions: Descriptive analytics – understanding the performance of the past (i.e., reporting) Predictive analytics – using data to anticipate how the future will look like (i.e., forecasting) Prescriptive analytics – suggesting a course of action to improve your business (i.e., optimizing)   As a manager, you should know what you want the data to do and recognize the five key benefits that business analytics yields:   1. Improved return-on-investment when compared to ‘pure analytics’ ‘Pure analytics’ means staring mining data without a specific business objective in mind. Such projects yield a high-risk of lack of results. Using business analytics, the business comes ahead of the data and guides the exploration process in a more consequent manner.   2. Superior robustness and interpretability of results when compared to ‘pure analytics’ As the business is guiding the overall process, from the problem framing to the validation of the solutions, teams are more likely to understand and use the results in the operation. Moreover, business sense should also steer away solutions that are only performing under strong assumptions.   3. More detailed decision-making when compared to just relying on business sense Without advanced analytics, managers often make decisions that work well on average, but fail to recognize the complexity of the business landscape. The power of data is connected to the possibility of tailoring decision to the different situations appropriately.   4. More alignment within your organization than with other approaches to decision making Business analytics strikes a good balance between grounding decisions on bottom-up evidence – data, while ensuring the appropriate business guidance. This equilibrium translates into team’s comfort as data brings the operational complexity with the appropriate business framing.   5. An opportunity to challenge your business beliefs With a fast-paced evolution of the expectations of the different echelons of the supply chain (e.g., suppliers and customers) it is ever more crucial to continuously challenge business beliefs that often lead to poor decision making. Business analytics, by allowing data exploration, is a very good instrument to cross-check managers intuitions.   Conclusion To reap the five key benefits that business analytics can bring, it is mandatory to have C-level support and the right team that blends business expertise and analytics prowess. The larger the number of stakeholders, who are able to manage both skills, the better.


As the COVID-19 pandemic is dramatically changing consumer behaviors worldwide, we look to the e-commerce business to understand how should retail companies navigate through the crisis. Back in 2003, with the outbreak of SARS, many companies in China started looking at online operations as a growth opportunity. Alibaba got an important breakthrough by expanding its operation from business-to-business to business-to-consumer, with the launch of Taobao. Taobao was a success and today is one of the largest e-commerce platforms. While e-commerce was still on its early stage on those days, today it is an important channel in most markets, but the question lingers: can the boom that e-commerce experienced in China in 2003 be replicated?   We analyze the new challenges and opportunities that online retailers are now facing, along two axes: Marketing & Sales, and Operations.   1st axis: Marketing & Sales The initial behavioral trends confirm that consumers will move massively to the online channel. This unfortunate crisis is an opportunity for e-commerce companies to scale. For many traditional retailers, it is actually a rare chance to finally shift volume to e-commerce, becoming less dependent on physical points of sale. To unlock such growth potential, marketing and sales departments need to be creative and dynamic, defining and deploying innovative campaigns. Customer engagement must be tactful and empathic: the priority is to build loyalty, while avoiding aggressive approaches that can lead to reputational damage, specially in these delicate times. Additionally, customer experience must be stellar: certain operational glitches must be avoided, such as out-of-stocks not properly marked or erroneous product descriptions. Despite this global opportunity for e-commerce, retailers that sell non-essential items may face a tougher situation, since consumers are likely to adopt a more conservative stance, delaying or avoiding such purchases. For such companies, the importance of putting assertive campaigns on the ground is even higher. Besides campaign design, impact tracking is another crucial function in the current context. Analytics play a decisive role in the live monitoring of both regular and promotional sales performance, enabling timely actions.   2nd axis: Operations From an operations point of view, increasing capacity is essential to face the expected rise in demand. For traditional retailers, moving capacity from physical retail to online operations is one of the internal opportunities. Increasing the efficiency of in-store picking, stock handling, packing and delivery operations should also be among the priorities of any operations manager. Last-mile deliveries can benefit particularly from increased and less restricted demand. Usually, most customers prefer short time windows after working hours, which hinders the delivery operation. However, the increase in customers’ availability due to the quarantine period adds flexibility to the retailer’s operations. Segmenting deliveries geographically allows to significantly cut travelling times between consecutive clients, reducing last-mile costs. Another possible step towards operational efficiency is to define a pricing policy adjusted to the customers’ willingness to wait. Offering or reducing the delivery fee to encourage orders with longer lead times, while increasing the fee for urgent deliveries, is a fair way to promote better planning. Crisis periods tend to reduce the resistance to change of both organizations and individuals. Therefore, online retailers should seize this chance of improving internal processes and capturing the above-mentioned opportunities. Still, there are also some challenges to tackle. It is paramount to assure the security of every employee and the full compliance with the updated health guidelines. Any failure related to health safety can seriously damage a retailers’ reputation. Additionally, the lack of capacity of outsourced services can interfere with the ability to face demand. E-commerce players should book in advance delivery capacity while notifying customers on real-time of potential delays. Click and collect is also an option to reduce pressure over outsourced services. Finally, quickly changing demand patterns also pose an operational challenge. Analytics is the key to successfully adapt the planning processes. For instance, a more reactive replenishment model might be the key to adjust inventory according to customers’ needs.   Takeaway All in all, these unusual times present significant challenges for e-commerce. Such is also the case for almost all other business sectors. Still, perhaps no other sector currently benefits from the same growth potential. Is it the assertiveness of decision-makers that will determine if e-commerce companies will embrace such window of opportunity, thus creating value for society by helping citizens stay safe while catering for their basic needs. Agility is paramount to navigate the uncertainty that companies face. Leaders must set autonomous multi-functional response units, from sales to operations, that can swiftly introduce the necessary changes to the business as usual planning process. Business and data analysts have a pivotal role in such units, bringing to the surface the new demand and supply patterns, and plotting different scenarios to support risk-aware strategic decisions.


In today’s uncertain and competitive world, it is mandatory to make better business decisions faster, and with better execution, in order to improve performance. Selecting one course of action (hopefully the best possible) from a set of available alternatives may turn into an intricate optimization problem. Despite the challenge of transforming insights and foresights into actionable smart and granular recommendations (e.g. optimize a production process or a marketing campaign, or design a robust supply network), the value derived from optimization (mathematical models and algorithms, and AI techniques) is enormous, as summarized in the next five tangible benefits:   1. It impacts company’s top-line and bottom-line and delivers fast return-on-investments Rules-of-thumb or simple decision rules underperform when used to solve complex decisions, and are useless on settings with high levels of uncertainty that may cause disruptions. Such empirical approaches deliver solutions far away from optimality. Prescriptive analytics impacts KPIs by expanding the solutions space (e.g. considering a wider planning horizon or integrating different functional areas of the organization), by comprehensively framing the problem in several manageable building blocks, differencing soft constraints (nice to meet) from hard constraints (violation is forbidden), and by guiding the search dynamically in function of the business goals to be optimized (thus exploring better tradeoffs between various conflicting criteria, such as customer satisfaction and operational efficiency).   2. It allows evaluating “what-if” scenarios to make more informed decisions The power of understanding the effect of decisions and of anticipating outcomes under changing assumptions and sources of uncertainty is critical in complex business challenges. By allowing the comparison of different solutions for alternative scenarios and inputs (e.g. demand, supply, pricing, costs and technological-related) in short running times, a prescriptive engine fosters new and innovative ways of solving the business problems, and of identifying business opportunities for growth and efficiency. Ultimately, identifying and better quantifying the risk, and devising mitigation strategies.   3. It streamlines the decision-making process for businesses A new way of problem solving and thinking in many organizations emerge, more holistic, moving from “gut-feel” decision-making to fact-based decisions. Actionable steps to be taken are supported by data. Prescriptive analytics brings more transparency to decision making (clarifying why a certain course of action is taken), blurring silos within the organizations and supporting effective, cross-functional interaction between teams. Enterprise-wide performance is thus favored, in contrast to improving performance in one area which may undermine the metrics in other process or functional areas.   4. It challenges internal status quo and crystalized ways of doing things Non-standardized, highly manual and time-consuming processes that are mostly not supported by analytical layers do not enable a forward-thinking approach. The necessary change of companies’ mindset regarding the use of optimization models and business decision support systems, requires more than just appropriate technology, people and processes. It requires proper change management which allows processes to get less dependent of the stakeholder involved (e.g. planner) and to avoid biased courses of action based on historical practices that may not have a proper fit for today’s volatile environment.   5. It blows up agility and responsiveness of the business Never before timely decisions were so critical. Unstructured decision-making processes on top of complex organizational structures hinder agility. By generating in short running times optimal solutions, prescriptive analytics helps us adjusting and improving companies’ response to rapidly changing conditions. Moreover, as data is being collected almost instantaneously, in many applications there is a need to reduce lead-times and potentially move to real-time decision making (e.g. order picking and order fulfillment related processes).   Conclusion Prescriptive analytics is the ultimate level of advanced analytics and is critical to gain competitive advantage now and in the future. More accurate and faster decisions, mitigating inherent risks, bring a transformational value into businesses.